Weird oscillation

[Wavebourn]

The amp oscillates on the rear side of one half - period of a sine wave. What is strange: it does not oscillate on the front half, but on the rear half, and does not start on the peak, but on the fall. What is strange also, an oscillation occurs when the signal frequency is below 100 Hz, and the lower is the frequency, the higher is an amplitude of oscillations. What can cause such a strange phenomenon?

Looks like a LC contour with very high Q, but where???

[Netlist]

Shorted inputs?
Wild guess... PS caps.

[Wavebourn]

Originally posted by Netlist
Shorted inputs?

No, inputs are fine. And they shorted by a 33 pF cap.

Wild guess... PS caps.

No, I tried adding 4 uF metal film to PS caps, though I can't try 800V source (no such cap), but I don't believe it's it, plus I have a simple voltage regulator for pre and screen grids (with additional RC for a pre). Indeed a voltage regulator needed a bigger grid stopper, now it is 1K, so I belive it is not it.

Parallel ground wires? I have no time now to switch it on and check the frequency, I have to prepare the stage for tonight's concert...

[wakibaki]

The description is a bit vague. It doesn't really sound like an oscillation certainly not in the sense of Barkhausen criterion sinewave oscillation. Can you post a pic preferably of a scope trace with timebase info?

Does the frequency vary when you tune the tone controls (I'm guessing that's what you mean) or just the amplitude? If you have a constant tone somewhere then tuning the circuit will have this effect on the amplitude.

How far back in the amp can you see the signal?

w

[Netlist]

Originally posted by Wavebourn
I have to prepare the stage for tonight's concert...

Not much time to fix things then.
I meant, did you short the inputs before measuring?

/Hugo

[Boris_The_Blade]

Kinda sounds like not enough power supply decoupling between the output stage and the driver stage.

[m6tt]

Some things to check / Strange oscillations I've seen:

-Microphonic tube is feeding back from speaker vibrations
-Input jack internal bypass tab is capacitively coupling to the input tab when a plug is in (if it's a 1/4th in. shorting jack)
-Acoustic feedback from amp --> instrument --> amp
-Bad speaker damping causing ringing
-Transformer causing ringing
-Weird bias?

[Wavebourn]

Originally posted by Netlist


Not much time to fix things then.
I meant, did you short the inputs before measuring?

No ground loops by design.
Wavetek functional generator is connected to input (2 wire power cord).
When input signal is above 100 Hz it looks like it has no time to oscillate (soft oscillation), when goes below 100 Hz an amplitude of oscillations is higher when input's signal frequency is lower. Starts on a fall of one half period. No oscillatioins on a rise.
I have no time to gather more info, tomorrow will continue.

Thanks guys!

[Wavebourn]

Originally posted by m6tt
Some things to check / Strange oscillations I've seen:

-Microphonic tube is feeding back from speaker vibrations
-Input jack internal bypass tab is capacitively coupling to the input tab when a plug is in (if it's a 1/4th in. shorting jack)
-Acoustic feedback from amp --> instrument --> amp
-Bad speaker damping causing ringing
-Transformer causing ringing
-Weird bias?

No acoustic feedback (also very high freqs!), but a transformer may ringing, but with such high Q?!
What do you mean about weird bias?
Dividers from -80 with pots to ground, 100K from dividers to coupling capos and 1K grid stoppers (GU50).

First tube: pentode half of 6F12P, Concertina on the triode half directly coupled to it's anode, from cathode of concertina goes to one grid of a diff stage (6N6P), from anode of Concertina goes to another grid through the 0.47uF cap, 2MOhm between grids.
Hmmm,,, Wirewound 4k7 in the tail... May be...?

[m6tt]

The Electricians Razor: Did it ever work?

The one-half wave oscillation made me think it might be something in the bias circuit, i.e. bias supply is somehow frequency affected, but only in a way that appears during a negative or positive swing...if it's p-p, that would seem like it would cancel. Could be in a preamp stage then. Have you monitored the bias current in various stages during this situation, does it fluctuate? Maybe the low note somehow affects operating conditions such that the gain of one of the tubes (pre-amp or power) actually increases? All of the oscillations in previously non-oscillating amps I've seen are either microphonic tubes, crappy speakers, lead dress or weird capacitive coupling in parts like jacks, switches and pots. I don't even own a scope, so I don't know much about what it *looks* like, but when I hear AF oscillations, they've always been for these reasons.

I'd try bypassing the pre-amp. The 60hz hum from touching a grid would be below the 100hz threshold. Does that oscillate with preamp detached? Also try other speakers to ensure that these aren't the problem (assuming you are scoping with a speaker attached). I've got a bit of experience, but I'm definitely often wrong. Just some ideas for quick checks I thought of given you need it for performance tonight. If you have some low value high voltage caps you could always band-aid the power supply and see if that cuts some ringing or oscillations out.